Our aim is to test the common assumption that aging in mammals involves a nonrandom decrease in neuron number and to consider the hypothesis that the decrease is preceded by changes in connections made and received by those neurons. The hypothesis derives from well-established data that at critical periods early in life, the number of neurons is adjusted downward by about 50% in many regions of the nervous system as a normal programmed event, and that axonal innervation of target cells also may be adjusted downward. Do such adjustments occur late in life and what controls them? Our analysis uses standard mice of the inbred C57BL/6 inbred strains, at ages from early adulthood to beyond 30 months of age, which is unusually old for mice. Complete serial section series through the whole brain are bing prepared by a new embedding and sectioning technique. We have developed computer-assisted quantitative methods applicable at light and electron microscopic levels for scoring the numbers and positions of neurons of defined classes, counting and classifying axons with statistical reliability in central tracts and peripheral serves, and measuring volumes and surface areas of brain regions and of individual cells. With the support of this grant, we are among the first to be able to measure fiber number and sizes automatically, directly from the electron microscope image, without photographic mediation. Further efforts will focus on the difficult micro-alignment and other problems inherent in three-dimensional quantitative reconstruction of synaptic organization at high magnification. We are concentrating overall on 3) numbers of cerebellar neurons in the major classes as a function of age, 2) the relation of retinal ganglion cell size, number, and distribution to the corresponding parameters in optic nerve, and 1) sampling and statistical methods for motor and sensory components in mixed peripheral nerves. Our objective is to establish methods and a normative baseline against which to test experimentally in the future, whether long-term neuronal survival is influenced by quantitative synaptic relationships at an earlier period of life.